Google deep learning and the local algo

[Margaret Ornsby]

Reading this:
https://medium.com/backchannel/google-search-will-be-your-next-brain-5207c26e4523

was fascinating in the first part, and then caused me to wonder a bit about the algos and the spam.

Is G is using deep learning to help sort out the search results? Seems so.
Could the issues we're seeing with spammy packs be the result of the deep learning system simply not getting the appropriate volume or good feedback about what spammy listings are?

If that's the case, it gives a very different perspective on how changes to the algo progress, and how G must have to choose to what to prioritise in the human intervention part to help the system learn what is spam and what isn't.

It also gives a different perspective on the "race against spammers" and how that plays out.

Also brings Baidu onto my radar in a way I hadn't considered before. If they're as into AI as Google, then does that make Baidu the biggest threat to G, not FB? FB might be the threat in terms of advertising revenue in the western world, but what about China with 18% of the world's population?

Raising this here as a discussion, not a problem. Just wondering what others know about this or have discovered in their research.

 

[Linda Buquet]

Thanks for sharing and starting the discussion Margaret!

Sounds fascinating. Can't wait to read!

Anyone have any other thoughts???

 

[JoshuaMackens]

Reading this:
https://medium.com/backchannel/google-search-will-be-your-next-brain-5207c26e4523

was fascinating in the first part, and then caused me to wonder a bit about the algos and the spam.

Is G is using deep learning to help sort out the search results? Seems so.
Could the issues we're seeing with spammy packs be the result of the deep learning system simply not getting the appropriate volume or good feedback about what spammy listings are?

If that's the case, it gives a very different perspective on how changes to the algo progress, and how G must have to choose to what to prioritise in the human intervention part to help the system learn what is spam and what isn't.

It also gives a different perspective on the "race against spammers" and how that plays out.

Also brings Baidu onto my radar in a way I hadn't considered before. If they're as into AI as Google, then does that make Baidu the biggest threat to G, not FB? FB might be the threat in terms of advertising revenue in the western world, but what about China with 18% of the world's population?

Raising this here as a discussion, not a problem. Just wondering what others know about this or have discovered in their research.

I don't think the neural nets have made their way into the search algorithm yet. Later on in the article they talk about application and it seems like while departments have access to them, nothing is being used officially.

But then again, you never know.

As far as Baidu being a threat, Google's aim, in my opinion, is much bigger than search. If Google X is any indication, Google's search engine is just their way to make money to fund their other projects. Google started out as a search company but their aims are much higher now. While AI can definitely help with search, Google has become much more than a search company and will only keep heading that way.

So, I don't really see Baidu as a threat to them. Baidu may be able to compete and own the Chinese sector which, to be sure, is or will be the biggest single country market for search (haven't seen latest stats/projections), but the Chinese population pales in comparison to the world population, which is Google's aim.

Also, I imagine if Google wanted China, they would go get it. They're the most powerful company on the planet and could hire/retain the talent if they wanted China bad enough. I just imagine it's a small matter on their plate currently. Again, all of this is conjecture without any stats of Chinese market share for both companies or projections on Chinese population and technological growth.

Awesome discussion!

 

[Dustybones]

Larry Page: ?Artificial intelligence would be the ultimate version of Google. So we have the ultimate search engine that would understand everything on the Web. It would understand exactly what you wanted, and it would give you the right thing. That?s obviously artificial intelligence, to be able to answer any question, basically, because almost everything is on the Web, right? We?re nowhere near doing that now. However, we can get incrementally closer to that, and that is basically what we work on. And that?s tremendously interesting from an intellectual standpoint. ?

(October 28, 2000)

 

[JoshuaMackens]

I was creeped out in the article when the "brain" gave the captions for those photos. It was unnerving.